THE FLOW OF WATER IN" CONCRETE PIPE. 37 



heavy brick oil gave a smooth surface to the interior of the pipe. 

 The concrete was tamped into the molds, and just before the initial 

 set was retamped by a finisher, settling in the forms about 3 inches. 



This line was designed by use of the Kutter formula, with a value 

 of n taken as 0.013. The maximum capacity of the whole siphon 

 was computed as 26.6 second-feet; but, whereas 29.39 second-feet 

 was sent through the pipe during observation No. 2, this discharge 

 practically filled the intake chamber and may be taken as the maxi- 

 mum capacity of the pipeline at the present time. 



Since there is a long stretch of wood-stave pipe beyond the con- 

 crete pipe, each slight change in the velocity through the latter made 

 a great change in the heights of the piezometer columns. A velocity 

 of 4.88 feet per second was the lowest that could be tested, as it 

 developed but a slight pressure head at the upper end of the reach. 

 The velocity for maximum capacity was 5.78 feet per second. Thus 

 the range of velocities feasible was rather short. Reference to Plate 

 VI shows that, while consistent, the arrangement of the observation 

 points indicates a very high exponent of V, namely, 3.18. Had the 

 full range of velocities — between, say 1 foot per second and 5.78 feet 

 per second — been possible of test, the points of observation might 

 have more nearly compared to a slope of 2. Otherwise the conditions 

 of experimentation appeared very good, and the writer is at a loss to 

 explain the behavior of the points. 



The friction factors, by the same token, vary, the coefficient C s 

 changing between 3.86 and 3.43, while from the appearance of the 

 pipe interior it should have been about 0.360 for all observations. 



Nos. 28 and 28b, Experiment S -54-55. — 36-inch monolithic pipe, 

 Clavey siphon, Oakdale irrigation district, California. — The Clavey 

 siphon of this district, built in 1912, is of unusual construction. 

 Water from an open channel lateral is conveyed 3,000 feet across a 

 swale in a 36-inch reinforced concrete pipe, built in place. The 

 outlet end is a few feet lower than the inlet. Thus far the construc- 

 tion is as usual. 



Near the bottom of the swale, a few hundred feet before reaching 

 the outlet, water from this pipe may be diverted through a booster 

 pump. The energy of the total flow in the 36-inch line is used to 

 pump a part of this same water through an additional 3,000 feet of 

 24-inch pipe to a point about 8 feet higher than the inlet of the 

 siphon.. 



Both parts of this siphon were constructed in the same manner. 

 Wood forms were used for inside and outside surfaces. This con- 

 struction usually gives longitudinal ridges in the concrete where each 

 crack between boards occurs and more or less irregularity in the 

 surface at the abutting sections. This is brought out in Plate I, 

 figure 1, and Plate II, figure 2. The mixture used was 1 part of 

 cement to 6 parts of gravel. The resulting concrete is not first class, 

 there being several leaks in the siphon. Most of these leaks occur at 

 the joints between abutting ends of pipe sections. Two reaches of 

 the 36-inch pipe were chosen for tests, No. 28Z> being 1,266 feet long 

 and No. 28 being 1,933.6 feet long, but including all of reach No. 286. 

 Two lengths were tested, as the inlet end of the pipe was not com- 

 pletely filled during observations 1 to 3, inclusive, because of the 

 small quantity of water, and the fact that the booster pump was not 

 in operation, the longer reach of pipe being tested when the booster 



